ARTICLE

https://doi.org/10.1038/s42003-020-01227-2 OPEN Nrf2 contributes to the gain of mice during space travel

Takafumi Suzuki 1,17, Akira Uruno1,2,17, Akane Yumoto3,17, Keiko Taguchi1,2,4, Mikiko Suzuki 5, Nobuhiko Harada6, Rie Ryoke7, Eriko Naganuma1, Nanae Osanai1, Aya Goto8, Hiromi Suda1, Ryan Browne7, Akihito Otsuki 2, Fumiki Katsuoka2,4, Michael Zorzi 9, Takahiro Yamazaki2, Daisuke Saigusa2, 1234567890():,; Seizo Koshiba2,4, Takashi Nakamura 10, Satoshi Fukumoto11, Hironobu Ikehata1, Keizo Nishikawa12, Norio Suzuki13, Ikuo Hirano2,8, Ritsuko Shimizu2,8, Tetsuya Oishi13, Hozumi Motohashi 14, Hirona Tsubouchi15, Risa Okada3,15, Takashi Kudo15, Michihiko Shimomura3, Thomas W. Kensler 16, Hiroyasu Mizuno3, ✉ ✉ Masaki Shirakawa3, Satoru Takahashi 15, Dai Shiba3 & Masayuki Yamamoto 1,2,4

Space flight produces an extreme environment with unique stressors, but little is known about how our body responds to these stresses. While there are many intractable limitations for in-flight space research, some can be overcome by utilizing knockout-disease model mice. Here, we report how deletion of Nrf2, a master regulator of stress defense pathways, affects the health of mice transported for a stay in the International Space Station (ISS). After 31 days in the ISS, all flight mice returned safely to Earth. Transcriptome and metabolome analyses revealed that the stresses of space travel evoked ageing-like changes of plasma metabolites and activated the Nrf2 signaling pathway. Especially, Nrf2 was found to be important for maintaining home- ostasis of white adipose tissues. This study opens approaches for future space research utilizing murine gene knockout-disease models, and provides insights into mitigating space-induced stresses that limit the further exploration of space by humans.

1 Department of Medical Biochemistry, Tohoku University Graduate School of Medicine, Sendai, . 2 Department of Integrative Genomics, Tohoku Medical Megabank Organization, Tohoku University, Sendai, Japan. 3 JEM Utilization Center, Human Spaceflight Technology Directorate, JAXA, Tsukuba, Japan. 4 The Advanced Research Center for Innovations in Next-Generation Medicine (INGEM), Tohoku University, Sendai, Japan. 5 Center for Radioisotope Sciences, Tohoku University Graduate School of Medicine, Sendai, Japan. 6 Institute for Animal Experimentation, Tohoku University Graduate School of Medicine, Sendai, Japan. 7 Department of Functional Brain Imaging, Institute of Development, Aging, and Cancer, Tohoku University, Sendai, Japan. 8 Department of Molecular Hematology, Tohoku University Graduate School of Medicine, Sendai, Japan. 9 Department of Molecular Biotechnology and Health Sciences, University of Turin, 10125 Torino, Italy. 10 Division of Molecular Pharmacology & Cell Biophysics, Tohoku University Graduate School of Dentistry, Sendai, Japan. 11 Division of Pediatric Dentistry, Tohoku University Graduate School of Dentistry, Sendai, Japan. 12 Immunology Frontier Research Center, Osaka University, Suita, Japan. 13 Division of Biology, Tohoku University Graduate School of Medicine, Sendai, Japan. 14 Department of Regulation, Institute of Development, Aging and Cancer, Tohoku University, Sendai, Japan. 15 Laboratory Animal Resource Center in Transborder Medical Research Center, and Department of Anatomy and Embryology, Faculty of Medicine, University of Tsukuba, Tsukuba, Japan. 16 Translational Research Program, Fred Hutchinson Cancer Research Center, Seattle, WA, USA. 17These authors contributed equally: Takafumi Suzuki, Akira Uruno, ✉ Akane Yumoto. email: [email protected]; [email protected]

COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio 1 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01227-2

uring space flight, astronauts experience harsh environ- To elucidate roles that Nrf2 plays in regulating adaptive Dments, including microgravity and high-dose cosmic responses to space stresses, in this study we conducted the one radiation, which affect the homeostasis of physiological month-long space travel of six Nrf2-KO mice and six WT mice. systems in our body1. To investigate how space flight affects health After a 31-day stay in the ISS in 2018, all of the flight mice returned of animals, space mouse experiments have been conducted safely to Earth. Using metabolomic as well as transcriptomic, his- exploiting the International Space Station (ISS) and other oppor- tological, morphometric, and behavioral methodologies, we found tunities2–10. However, it has been difficult to attain live return of that Nrf2 signaling was indeed activated by space stresses in various mice from space or to even realize “space travel” of mice. For tissues. Space stress and Nrf2-deficiency brought about changes in instance, the Italian mice drawer system housed six male mice gene expression and plasma metabolite profiles independently for individually, but more than half of the animals died during habi- the most part, but cooperatively in certain situations. In particular, tation in space8–10. Many of the other preceding space mouse Nrf2 is important for the weight-gain of space mice and main- projects did not attempt live return of the mice from space. tenance of white adipose tissue homeostasis in response to the To achieve live-return of space mice, the Japan Aerospace stresses of month-long space travel. Exploration Agency (JAXA) recently established a fully-equipped mouse experimental system for space flight11. It comprises mouse habitat cage units (HCU), transportation cage units (TCU), and a Results centrifuge-equipped biological experiment facility (CBEF). These Outline of MHU-3 mouse project utilizing Nrf2-KO mice.To apparatuses in combination can house mice individually and study contributions of Nrf2 to the protection of mice against the realize artificial gravity experiments in space. In the first and stresses of travel to and from space, and maintenance of home- second missions utilizing this equipment, referred to as Mouse ostasis during their space stay, we have conducted the MHU-3 Habitat Unit-1 and -2 (MHU-1 and MHU-2 conducted in 2016 project. Male Nrf2-KO mice16 and WT mice were bred and and 2017, respectively), 12 wild-type (WT) male mice were suc- selected for space travel. For this purpose, 60 WT and 60 Nrf2- cessfully launched each time, and all the mice returned safely to KO mice at 8-week-old were delivered to the Kennedy Space Earth after approximately 1 month stays in the ISS11,12. One Center (KSC) 3 weeks prior to launch. These mice were accli- salient finding in these missions is that ageing phenotypes, such matized to individual housing cages. After acclimation, we as reduction of bone density and muscle mass, were markedly selected 12 mice for flight to the ISS based upon their body accelerated during space flight; notably, these phenotypes could weight, levels of food consumption and water intake and the be prevented by housing in space with artificial gravity. phenotypic absence of a hepatic shunt (see “Methods”). It has been shown that various environmental stresses, SpaceX Falcon 14 rocket (SpX14) containing the mice in the including oxidative and toxic chemical (often electrophilic) TCU within the Dragon capsule was launched on April 2, 2018 stresses, activate Nrf2 and downstream signaling pathways13. (GMT) from KSC (Fig. 1a). After arrival at the ISS, mice were Nrf2 is the master transcription factor mitigating oxidative stress. relocated to the HCU by the crew. We used an HCU that Nrf2 induction is well-known to prevent various diseases, accommodates one mouse per cage11,12. Before mice were including cancer, diabetes, and inflammation13. As cosmic returned to Earth, they were transferred back into the TCU and radiation induces oxidative stress1 and mechanical stresses can loaded into the Dragon capsule which subsequently splashed also induce Nrf2 activity14, we hypothesized that space stresses down in the Pacific Ocean offshore from Southern California on may activate the Nrf2 signaling pathway allowing, in turn, for May 5. The TCU was unloaded from the capsule and transported Nrf2 to play important roles in regulating adaptive responses to to Long Beach Port on May 7. All mice were alive upon return these space stresses. Although one preliminary analysis exploiting and were then euthanized and dissected at a laboratory to collect astronaut’s hair samples during space flights showed that tissues after a general health assessment and a series of behavioral expression of NRF2 was decreased in their hair roots during space tests. flight15, the experimental design harbored limitations. Therefore, A ground control (GC) experiment precisely simulated the we decided to address further studies that would provide direct space experiment was conducted at JAXA Tsukuba in Japan from lines of evidence on this point. September 17 to October 20, 2018. Six WT and six Nrf2-KO mice In order to test the hypothesis that Nrf2 contributes to the were individually housed in the same units as the flight adaptive maintenance of animal homeostasis in response to the experiment (FL). Both the TCU and HCU were placed in an stresses of space flight, it was critical to establish space travel of air-conditioned room with a 12-h light/dark cycle. Fan-generated Nrf2 gene-knockout model mice16. To do this, however, we airflow (0.2 m/s) inside the HCU maintained the same conditions needed to overcome various challenges, both technical and reg- as in the FL setting. ulatory, as space travel of gene-modified disease-model mouse During the onboard habitation, the health conditions of each lines coupled with a need for live-return had never been con- mouse was monitored daily by veterinarians on the ground via ducted. To this end, we decided to utilize the MHU technologies downlinked videos (Yumoto et al., in preparation). Representative and proposed in 2016 an experiment for JAXA to send Nrf2- images of all 12 mice in the HCU and their onboard movies are knockout (Nrf2-KO) mice to space. shown in Fig. 1b and Supplementary movie 1, respectively. In this regard, metabolite of body fluids are During the flight mission, , and humidity were well considered as quantitative traits that can describe a real-time controlled, and concentrations of carbon dioxide and ammonia snapshot of physiological state of animals17,18. However, little is were maintained at low levels11. The absorbed dose rate of understood about the response of plasma metabolites to the space radiation was 0.29 mGy/day during the flight. stress. Due to the constraints of space experiments, there had A new device was developed to collect peripheral blood from been no attempt that focused on changes of metabolites in the tail in space (Fig. 1c). We obtained approximately 40-μL onboard and post-flight rodent blood samples. Recently, meta- blood from each mouse with minimal hemolysis (Fig. 1d). Blood bolome technologies based on mass-spectrometry have achieved collections from tail veins were performed the same way as in the the sensitivity required for analyses of very small amounts of space before launch and after return to Earth, as well as in the GC blood19. Moreover, metabolite profiling by nuclear magnetic experiment. resonance (NMR) has also become a precise and reproducible Initial inspection of mice returned to Earth showed loss of method for biomarker discovery17,18. balance and enfeebled muscle power (Supplementary Movie 2).

2 COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01227-2 ARTICLE

Fig. 1 The JAXA mouse project using Nrf2-KO mice. a Overview of the MHU-3 study. Six wild-type mice (WT) and six Nrf2-KO mice (KO) in the C57BL/ 6J background were launched by SpX14 from the Kennedy Space Center in Florida on April 2, 2018. After 31 days of habitation on the ISS, the mice splashed down in Pacific Ocean near Southern California on May 5. The mice were transported to a laboratory for behavioral observations and the other analyses two days later. b Representative images of onboard habitation for WT and Nrf2-KO mice. c Overview of the blood collection procedure. All steps are carried out while in orbit: (1) transfer each mouse to a restraining device (covered by a disposable heat pad prior to transfer); (2) use a tail clipper to cut off a 1 mm tip of the mouse tail; (3) collection of blood into a capillary tube inserted into a centrifuge tube, followed by centrifugation; and (4) hemostasis of the tail by applying using a hemostat. d Representative pictures of blood samples of flight and ground controls. Arrowheads indicate the plasma fraction with the absence of hemolysis.

The Nrf2-KO flight mouse No. 4 (FL04) developed severe brown adipose tissue (iBAT), thymus (Thy), kidney (Kid), and intestinal hemorrhage during the return flight for an unknown brain cerebrum (Cbr) (Fig. 2a). We selected 26 that encode reason. Data from FL04 were deemed outliers for many for detoxication and antioxidative responses that are well- measurements; therefore, data for FL04 were omitted for most known Nrf2 target genes20. The expression of these typical Nrf2 of the analyses. target genes was upregulated widely in various tissues of the FL_WT mice compared with those in GC_WT mouse tissues Space stresses activate the Nrf2 signaling pathway.Toexamine (Fig. 2b). whether space stresses activated Nrf2 and downstream pathways, In contrast, expression of these 26 selected genes were we conducted a wide-ranging RNA-sequence analysis. For this markedly lower in tissues of the GC and FL Nrf2-KO mice purpose, we dissected the space mice soon after their return to (Fig. 2b), indicating that the upregulation of these genes was Earth and prepared RNA samples from many tissues, including attributable to the functional presence of Nrf2 signaling. By temporal bone (TpB), mandible bone (MdB), spleen (Spl), liver comparing FL_KO with GC_KO, we examined all the gene (Liv), epididymal white adipose tissue (eWAT), inter-scapular expression changes induced by space flight in the WT mice. The

COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio 3 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01227-2

a WT KO TpB MdB GC Spl Liv RNA-sequence eWAT iBAT analysis FL Thy Kid Cbr b GC_WT FL_WT GC_KO FL_KO TpB MdB Spl Liv eWAT iBAT Thy Kid Cbr TpB MdB Spl Liv eWAT iBAT Thy Kid Cbr TpB MdB Spl Liv eWAT iBAT Thy Kid Cbr TpB Abcc2 MdB Akr1b10 Aldh2 Spl Fth1 Liv Ftl1 Gclc eWAT Gclm iBAT Gpx2 Gsr Thy Gss Kid Gsta2 Cbr Gsta3 Gsta4 Set as 0 Gstm1 Gstm3 Gstm4 Gstp1 Hmox1 Mafg Me1 Nqo1 Osgin1 Srxn1 Taldo1 Tkt Txnrd1 Log2FC vs GC_WT c 2 10 1 2 d Nqo1 Hmox1 Space stress p=0.01 p=0.05 30 12 P<0.01 P<0.01 Microgravity 25 10 Cosmic radiation

20 8 Nrf2 activation 15 6 TPM TPM

10 4 Up-regulation of 5 2 cytoprotective genes

0 0 !"#$%&GC FL GC FL GC FL GC FL Stress defense WT KO WT KO

Fig. 2 Space stresses activate the Nrf2 pathway. a RNA-sequence analyses of tissues from WT and Nrf2-KO mice in GC and FL groups. b Heatmap of expression levels of representative Nrf2-target genes in temporal bone (TpB), mandible bone (MdB), spleen (Spl), liver (Liv), epididymal white adipose tissue (eWAT), inter-scapular brown adipose tissue (iBAT), thymus (Thy), kidney (Kid), and brain cerebrum (Cbr) from WT and Nrf2-KO mice in GC and FL groups. c Expression levels of Nqo1 and Hmox1 genes in thymus from WT and Nrf2-KO mice in GC and FL groups. Data are presented as mean, and dots represent individual animals. n = 6 for GC_WT, FL_WT and GC_KO, and n = 5 for FL_KO, one-way ANOVA with Tukey–Kramer test. d Model for space stress response system by Nrf2 signaling. results of gene set enrichment analysis (GSEA) suggest that a of the extent of gene expression reduction revealed that the number of pathways are affected by space flight (Supplementary response was weaker in FL_KO mice than in GC_KO mice. We Table 1). The heatmap analyses revealed that space flight-induced surmise that that activation of other stress response pathway(s) by changes of gene expression in various tissues could be classified the strong space stresses might alter the magnitude of repression into a Nrf2-dependent group and a Nrf2-independent group (or loss of constitutive expression) elicited by the Nrf2 gene (Supplementary Fig. 1a–e). Of note, Nrf2-dependent space- knockout in the space flight mouse tissues. However, it should be induced genes include typical Nrf2 target genes. Closer inspection noted that the Nrf2 pathway contribution is the strongest among

4 COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01227-2 ARTICLE the regulatory pathways for the expression of these genes, as to examine changes in plasma metabolites associated with ageing. reductions of expression were substantial in both GC and FL Herein, we addressed whether ageing-related changes of meta- Nrf2-KO mice compared with both GC and FL WT mice. bolites were accelerated by the deficiency of Nrf2-regulated We extended these heatmaps to direct measurements, and cytoprotective systems. To this end, we collected blood plasma observed significant induction of transcriptional expression of from the mouse inferior vena cava soon after the return of the Nqo1 and Hmox1 in thymus from FL_WT mice (Fig. 2c). In very mice and conducted NMR-based metabolome analyses (Fig. 4a). good agreement with the heatmaps, in some of the tissues of Whereas PCA of the metabolome results did not show strong FL_KO mice, the magnitudes of lower expression were separations compared to that of transcriptome, closer inspection dampened, but still the space-flight induced increase of Nrf2- indicated that PC1 separated FL_KO vs. FL_WT (Fig. 4b), sug- target gene transcripts were largely canceled in the knockout gesting that both space flight and Nrf2-deficiency contributed to mice. These results unequivocally demonstrate that Nrf2 activity changes in plasma metabolites. is indeed induced during space flight and enhances the expression We then searched for plasma metabolites, which changed only by of cytoprotective genes, strongly arguing that our body exploits space flight or changed by both space flight and Nrf2-deficiency. Of the Nrf2 signaling pathway to counteract space stresses (Fig. 2d). all 40 metabolites examined by NMR-based metabolome analyses (Fig. 4c–e, Supplementary Figs. 3 and 4), we identified three fl fi fl metabolites of interest; i.e., glycerol, glycine, and succinate Space ight and Nrf2-de ciency in uenced gene expression.To – fl further clarify contributions of Nrf2 to gene expression patterns (Fig. 4c e). Plasma glycerol levels were increased by ight in both WT and Nrf2-KO mice compared to respective GC mice and with of the space mouse, we performed principal component analyses fl fi (PCA) of the transcriptomic results. We utilized all identified little in uence of Nrf2-de ciency (Fig. 4c). Plasma levels of glycine transcripts for this analysis. Results of the PCA revealed that (Fig. 4d) and succinate (Fig. 4e) in FL_WT mice were much lower space travel and Nrf2-deficiency both influenced gene expression than GC_WT mice. Glycine and succinate levels in GC_KO mice profiles, resulting in four distinct patterns. In thymus and eWAT, were also much lower than those in GC_WT mice, and levels did PC1 separated FL vs. GC, while PC2 separated Nrf2-KO vs. WT not decrease further with space travel. We additionally found that (Fig. 3a, b). By contrast, in liver and spleen, PC1 separated Nrf2- plasma levels of glutamine, carnitine and formate were changed significantly by the space flight (Supplementary Fig. 3). These KO vs. WT, while PC2 separated FL vs. GC (Fig. 3c, d). These fi results delineate the first and second patterns in which both space results imply that Nrf2-de ciency itself evoked similar changes to stresses and Nrf2-deficiency strongly and independently elicited those provoked by space stresses. specific changes in gene-expression profiles. An intriguing hypothesis here was that these changes in A third pattern was found in iBAT in which PC1 separated FL metabolites recapitulated the changes in metabolites accompanied vs. GC, while PC2 did not separate the other groups at all with ageing of humans. To address this hypothesis, we exploited (Fig. 3e), indicating that space travel influenced the gene human metabolome data accumulated in the Tohoku Medical fi Megabank (ToMMo) project22. We examined these metabolites in expression in this tissue while Nrf2-de ciency did not. We found – – a fourth pattern in Cbr. Somewhat to our surprise, there was no plasma from young age (20 40 years old) and old age (60 80 years clear separation of gene expression patterns by PCA in Cbr old) participants in the population-based prospective cohort studies fi of ToMMo. We found that plasma glycerol level was increased in (Fig. 3f), indicating that neither space stresses nor Nrf2-de ciency – affected gene expression in this tissue when analyzed en bloc. We 60 80 age group (Fig. 4f), showing very good accord with the space fl mouse results. Plasma levels of glycine (Fig. 4g) and succinate surmise that this pattern re ected the cell heterogeneity of Cbr. – – Collectively, these four PCA patterns of transcriptome analyses (Fig. 4h) were lower in 60 80 age group than in 20 40 age group. fl These changes again showed very good agreement with those in the demonstrated that, in most of the cases, space stresses in uenced fl gene expression differently than Nrf2-deficiency. ight mice. Importantly, the latter two metabolites were also decreased in Nrf2-KO mice, supporting the notion that Nrf2 is important to decelerate the ageing of animals. In contrast, while No acceleration of bone/muscle degeneration in Nrf2-KO mice plasma levels of glutamine, carnitine and formate were changed in space. Recent MHU1 and MHU2 reports revealed that ageing significantly by the space flight (Supplementary Fig. 3), the levels phenotypes, such as reduction of bone density and muscle mass, either changed moderately (glutamine and carnitine) or to a were accelerated during the month-long space flights11,21. Our fi reversed-direction in the human ageing analysis (Supplementary results nicely con rmed these observations. Micro-computed- Fig. 5). These results thus suggest that space stress induces ageing- tomography images clearly showed decreased bone density of like changes within a subset of plasma metabolites, and that Nrf2- FL_WT and FL_KO mouse bones compared with GC_WT and deficiency also provokes similar ageing-like changes of plasma GC_KO mouse bones (Supplementary Fig. 2a, b). Against our metabolites (Fig. 4i). expectation, Nrf2-gene knockout did not accelerate the decrease In addition, we also examined whether space flight induced of bone mineral density (BMD) during space travel. Muscle mass ageing-like changes in gene expression. GSEA using the gene set of soleus (Supplementary Fig. 2c) and gastrocnemius muscles of aged mice (Enrichr) revealed that space-induced changes of (Supplementary Fig. 2d) also showed substantial decreases during gene expression were enriched in ageing changes in liver, TpB, space travel, and again loss-of-Nrf2 did not accelerate nor BAT, and WAT (Supplementary Fig. 6). These results demon- decelerate these declines. These results indicate that Nrf2- fi fl strate that space stress induced ageing-like changes in metabolites de ciency did not in uence substantially the progression of and gene expression. these ageing phenotypes in space. We envisage that the space stress-originated phenotypes of bone and muscle in the mice are very strong, whereas baseline expression levels and impact of Lack of body-weight-gain in Nrf2-KO mice during space flight. Nrf2 are low, so that any possible Nrf2 contribution could not be Upon return to Earth, we examined the overall health status of FL visible in this context. and GC mice, conducted behavioral examinations, and, after necropsy, histological examinations of multiple tissues of the Space flight induces ageing-like changes. The observation that mice. One of the most obvious phenotypes we found in these there was no apparent acceleration of bone and muscle changes examinations was the lack of body-weight-gain specifically in during space flight of Nrf2-KO mice compared to FL_WT led us Nrf2-KO mice during the space flight (Fig. 5a). Since we launched

COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio 5 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01227-2

a b Thymus eWAT 10 GC_WT FL_KO GC_KO 5 5 FL_WT

0

PC2 0 PC2 FL_KO FL_WT GC_WT -5 -5 GC_KO

-10 -5 0 5 -10 -50 5 10 PC1 PC1 c d Liver Spleen 10 FL_WT FL_KO 5 FL_KO

5 GC_WT 0 PC2 GC_WT 0 FL_WT GC_KO -5 -5 GC_KO -10 -10 -5 0 5 10 -5 0 510 PC1 PC1

e 10 f 6 iBAT Cbr FL_WT GC_KO 3 0

GC_WT 0 GC_WT PC2 PC2 PC2 FL_WT GC_KO -10 -3 FL_KO FL_KO -6 -10 0 10 -5 05 10 PC1 PC1

Fig. 3 RNA-sequence analyses of flight and ground control mice. a–f Plots for PCA applied to RNA-sequencing data of thymus (a), eWAT (b), liver (c), spleen (d), iBAT (e), or Cbr (f) from WT and Nrf2-KO mice in GC and FL groups. Dots represent individual animals. n = 6 for GC_WT, FL_WT and GC_KO, and n = 5 for FL_KO.

11-week-old mice, the mice were still gaining weight. In fact, proportional to body weight did not change much by either space FL_WT mice as well as both GC_WT and GC_KO mice gained flight or Nrf2-deficiency or both (Fig. 5d). body-weight almost to the same extent. We designed an apparatus to monitor food-intake and water- We then examined of organs and tissues of these mice, consumption of mice while in space (Fig. 5e). Importantly, there including eWAT, iBAT, liver, spleen, lung, thymus, heart, testis, and were no significant differences in food-intake and water- kidney (Fig. 5b–d and Supplementary Fig. 7). We found that eWAT consumption between WT and Nrf2-KO mice, both in-flight and weight, as percent of body weight, was significantly increased by on the ground (Fig. 5f, g). These results demonstrate that during space fight, but this increase was canceled largely in the FL_KO space flight Nrf2-deficiency results in the reduction of body weight mice (Fig. 5b). Importantly, this decrease of eWAT weight did not without changing food-intake and water-consumption. Available occur in GC_KO mice. Similarly, weights of iBAT were significantly lines of evidence suggest that this may be linked to the diminished increased in FL_WT mice. However, in contrast to the situation of weight-gain of abdominal adipose tissues. eWAT, the increase was not canceled in the FL_KO mice (Fig. 5c). Therefore, to obtain further insight into the effects of space flight Showing stark contrast to these two adipose tissues, the liver weight on and glucose status in animal body, we analyzed the small

6 COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01227-2 ARTICLE

a WT KO b 100 FL_KO GC GC_WT FL

0 PC2

Blood collection FL_WT from inferior vena cava

GC_KO NMR -100 Metabolome analysis -200 0 200 PC1

cdGlycerol (µM) Glycine (µM) e Succinate (µM) p=0.05 p=0.01 250 p<0.01 12 p<0.01 p<0.01 p<0.01 500 200 10

8 150 400 6 100 4 300 50 2

0 200 0 !'!! "'!!GCFL #'!! $'!!GCFL %'!! &'!! !"#$%&GC FL GC FL !'!! "'!!GC #'!! FL $'!!GC %'!! FL &'!! WT KO WT KO WT KO

f Glycerol (µM) g Glycine (µM) h Succinate (µM) P=1.28e-08 P=1.48e-05 P=0.0246 200

150 300 10

100 200 5 50

100 20-40 60-80 20-40 60-80 20-40 60-80 age age age age age age

i Ageing-associated Space stress Glycerol metabolites Microgravity/Cosmic radiation Glycine Nrf2 deficiency Succinate

Fig. 4 Space flight induces ageing-like changes of plasma metabolites. a NMR-based metabolome analyses of plasma collected from the inferior vena cava of WT and Nrf2-KO mice in GC and FL groups. b Plots for PCA applied to NMR-based metabolome analyses of plasma from WT and Nrf2-KO mice in GC and FL groups. Dots represent individual animals. c–e Plasma levels of glycerol (c), glycine (d), and succinate (e) in WT and Nrf2-KO mice in GC and FL groups. Data are presented as mean, and dots represent individual animals. n = 6 for GC_WT, FL_WT and GC_KO, and n = 5 for FL_KO. One-way ANOVA with Tukey–Kramer test. f–h Age-distribution of plasma levels of glycerol (f), glycine (g), and succinate (h) in a human cohort of the ToMMo study (data replotted from ref. 22). Center line, median; box limits, upper and lower quartiles; whiskers, 1.5× interquartile range; points, outliers. P values were calculated with Wilcoxon–Mann–Whitney test. n = 545 (20–40 age) and 955 (60–80 age) for (f, g). n = 543 (20–40 age) and 948 (60–80 age) for (h). i Identification of ageing-associated metabolites from space mouse experiment and the ToMMo human cohort study. amounts of blood plasma collected from the mouse tail during flight total cholesterol ester (CE) levels between WT and Nrf2-KO mice (after 18 days launch, L + 18) and 2 days after return to Earth (R + during flight (L + 18) (Fig. 5i), there was an elevation of total CE 2). To our best knowledge, this is the first analysis of blood obtained levels in FL_WT mice, but not Nrf2-KO mice after return to Earth from mice in space. We conducted mass-spectrometry metabolome (R + 2), where levels mirrored those of GC levels (Fig. 5j). While we analyses of the plasma (Fig. 5h). While there was no difference in do not have solid explanation for the increase of total CE level only

COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio 7 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01227-2

in the WT mice after return to Earth (R + 2), it plausible that the to the total CE level, the total TG level was significantly decreased in decrease of eWAT in FL_KO mice observed after return to Earth FL_WT mice upon return to Earth (R + 2) compared with GC_WT (R + 2) (Fig. 5b) might be associated with the unchanged total CE mice of R + 2(Fig.5l). This tendency was similar in GC and FL level in the plasma of these mice. Nrf2-KO mice after return to Earth (R + 2) (Fig. 5l). Considering Similarly, total (TG) levels did not change much the elevated glycerol level in plasma (Fig. 4c), lipolysis from TG to between FL_WT and GC_WT mice, and also between FL_KO and glycerol was likely accelerated in flight mice in both WT and Nrf2- GC_KO mice during flight (L + 18) (Fig. 5k). However, in contrast KO mice. In addition, our NMR metabolome analysis revealed that

8 COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01227-2 ARTICLE

Fig. 5 Nrf2 deficiency affects body weight gain and abdominal white adipose tissue. a Body weight change of WT and Nrf2-KO mice from before launch to after flight in GC and FL groups. Note that Nrf2-KO mice failed to gain body weight during space flight unlike WT mice. n = 6 for GC_WT, FL_WT and GC_KO, and n = 5 for FL_KO. b–d Body weight-normalized masses of eWAT (b), iBAT (c), and liver (d) from WT and Nrf2-KO mice in GC and FL groups. n = 6 for GC_WT, FL_WT and GC_KO, and n = 5 for FL_KO. Note that eWAT increased significantly in WT mice during space flight, but the increase was smaller in Nrf2-KO mice. e Food intake was estimated by remaining food scale of food bar cartridge. f, g Food intake (f) and water consumption (g)ofWT and Nrf2-KO mice in GC and FL groups. n = 6 for GC_WT, FL_WT and GC_KO, and n = 5 for FL_KO. h Timeline for mass spectrometry-based metabolome analyses of plasma from tail blood of WT and Nrf2-KO mice in GC and FL groups. i–l Plasma levels of total cholesterol ester (CE) (i, j) and total triglyceride (TG) (k, l) and in WT and Nrf2-KO mice in GC and FL groups during flight (L + 18) (i, k) and after return (R + 2) (j, l). Data are presented as mean, and dots represent individual animals. n = 4 (GC_WT), n = 5 (FL_WT), n = 4 (GC_KO), and n = 5 (FL_KO) for (i). n = 5 (GC_WT), n = 5 (FL_WT), n = 5 (GC_KO), and n = 5 (FL_KO) for (j). n = 6 (GC_WT), n = 5 (FL_WT), n = 6 (GC_KO), and n = 5 (FL_KO) for (k). n = 5 (GC_WT), n = 6 (FL_WT), n = 6 (GC_KO), and n = 5 (FL_KO) for (l). One-way ANOVA with Tukey–Kramer test.

there was no significant change in plasma glucose levels genes involved in the respiratory chain (Fig. 7a) and β- (Supplementary Fig. 5). These results thus demonstrate that the oxidation (Fig. 7b) were markedly reduced in eWAT from the space travel affects lipid in mice, and the changes in flight mice of both genotypes compared with expression levels in certain aspects of were either reversed or eWAT from GC WT mice. In addition, similar but milder exacerbated by the loss-of-Nrf2 function. changes than those in the flight mice were observed in the GC_KO mice (Fig. 7a, b), indicating that space stresses induce metabolic impairment in eWAT, which is qualitatively similar, Nrf2 is critical for maintenance of WAT homeostasis in space. but much more severe, than those observed with Nrf2-deficiency We then extended our mouse analyses to the histology of mice on Earth. These results suggest that these reductions of abdominal WAT. We observed that lipid droplet size of eWAT in fi mitochondrial activity might result in the enlargement of adi- GC Nrf2-KO mice was signi cantly larger than that of GC_WT pocytes in FL_WT and GC_KO mice. mice (Fig. 6a) despite noting that weights of eWAT were com- Further transcriptome analysis revealed that expression levels of parable between these two GC groups (Fig. 5b). Surprisingly, 23 fl diabetes-related chemokine genes were increased in eWAT from space ight gave rise to a marked increase of lipid droplet size in flight mice of both genotypes compared to GC mice (Fig. 7c). WT mice and to some extent in Nrf2-KO mice (Fig. 6a). We Consistent with this observation, expression levels of marker genes measured sizes of lipid droplets of all mice, and found that these fl for macrophage (Cd68, Lgals3,andAdgre1) and angiogenesis (Kdr changes during the space ight were quite reproducible (Fig. 6b). and Pecam1) were increased in eWAT from flight mice of both We also measured the distribution of lipid droplet sizes and fi fl genotypes compared to GC mice (Fig. 7d). Since angiogenesis con rmed that space ight provoked increases of lipid droplet sustains inflammation by delivering oxygen and nutrients for size in both WT and Nrf2-KO mice (Fig. 6c, solid lines) com- inflammatory cells, and inflammation in turn can cause insulin pared with two groups of GC mice (dotted lines). resistance24, these results suggest that space stresses might induce We calculated adipose cell number of eWAT utilizing these adipose tissue growth by means of inflammation and angiogenesis. data, and found that the number of adipose cells in Nrf2-KO fi fi In order to gain insight as to how Nrf2 de ciency on Earth leads mouse eWAT was signi cantly lower than that in WT mouse to the decrease in adipocyte number, we searched for changes in eWAT (Fig. 6d), indicating that eWAT weight of Nrf2-KO mice gene expression in eWAT from GC_KO mice. We found that on the ground is maintained by a complementary increase in expression levels of many PPARγ-target genes25 were down- droplet size. Intriguingly, lipid droplet size of FL_WT mouse fl regulated in eWAT from GC_KO mice compared to GC_WT mice eWAT became much larger following space ight than that of the (Fig. 7e). Since PPARγ is important for adipocyte differentiation25, GC mice. However, the lipid droplet size in eWAT of FL_KO this downregulation of PPARγ in GC_KO mice might affect mice was almost comparable with that of FL_WT mice (Fig. 6b), adipocyte differentiation, provoke reduction of adipocyte numbers indicating that eWAT of the Nrf2-KO mouse did not become fl and, in turn, give rise to compensatory enlargement of the larger during space ight (Fig. 6e). Taken together, we propose adipocytes, thereby sustaining adipose mass of eWAT on Earth that Nrf2 plays important roles in the maintenance of abdominal (Fig. 7f). The transcriptome analysis also revealed that expression of adipose tissue homeostasis, but that space stresses markedly affect a number of PPARγ-target genes were changed in the eWAT of this homeostasis regardless of the presence of Nrf2. flight mice, showing similar profiles irrespective of the Nrf2 In an associated observation, lipid droplet size and thickness of genotypes (Fig. 7e). The profile of flight mice showed remarkable subcutaneous fat was found to be increased by space flight, fi differences from those of Nrf2 KO mice on Earth. These results thus although there was no signi cant difference between WT and indicate that space stresses elicit much stronger influences on the Nrf2-KO mice (Supplementary Fig. 8). Furthermore, numbers of PPARγ-target gene expression, along with the respiratory chain, hepatic Oil Red O-positive lipid droplets were increased by space fatty acid β-oxidation, chemokine and macrophage-/angiogenesis- flight (Supplementary Fig. 9). Tissue weights and lipid droplet fl related gene expressions, than the Nrf2 knockout does. We surmise sizes of iBAT also increased through space ight in both WT and that these changes in the gene expression profiles are, at least in Nrf2-KO mice (Fig. 5c and Supplementary Fig. 10). These wide- part, responsible for the marked enlargement of the adipose tissues ranging observations demonstrate that adipose tissues in the during the space travel. whole body became larger in response to space stresses, supporting the contention that Nrf2 is critical for the main- Discussion tenance of homeostasis of abdominal WAT. Nrf2 is the key regulator of the adaptive response against various environmental and endogenous stresses13. Since oxidative stress Space flight and Nrf2-KO induce metabolic impairment in activates the Nrf2 signaling pathway through stress-sensing eWAT. To elucidate how space flight and Nrf2-deficiency induce mediated by the Nrf2 chaperone Keap1 (Kelch-like ECH- perturbations in eWAT, we examined the transcriptome. Ana- associated 1)26, cosmic radiation and/or microgravity lyses of eWAT-derived RNA revealed that mRNAs coding for during space travel may well activate Nrf2 through the generation

COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio 9 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01227-2

Fig. 6 Nrf2 is essential for maintaining eWAT mass in space. a Histological images of eWAT from WT and Nrf2-KO mice in GC and FL groups. b, c Average lipid droplet size (b) and lipid droplet size distributions (c) of eWAT from WT and Nrf2-KO mice in GC and FL groups. Data are presented as mean, and dots represent individual animals. n = 6 for GC_WT, FL_WT and GC_KO, and n = 5 for FL_KO. d Calculated adipose cell number of eWAT from WT and Nrf2-KO mice in GC and FL groups. Data are presented as mean, and dots represent individual animals. n = 6 for GC_WT, FL_WT and GC_KO, and n = 5 for FL_KO. e Model for adipose cells in eWAT from WT and Nrf2-KO mice in GC and FL groups. Data are presented as mean, and dots represent individual animals. One-way ANOVA with Tukey–Kramer test.

10 COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01227-2 ARTICLE

a Respiratorychain b Fatty acid e PPAR relatedgenes -oxidation target genes WT KO relatedgene WT KO GC FL GC FL WT KO GC FL GC FL mt-Nd6 1.2 Ces1f Uqcrc2 GC FL GC FL 1.2 Hcar1 mt-Nd5 Acacb 1.5 Slc1a3 Ndufa1 Echdc1 mt-Nd4 Ivd Fam13a 1.0 Sdha Echs1 1.0 Angptl4 1.0 Ndufa10 Cpt2 Adipoq mt-Co1 Hibch Pnpla2 Acat2 Fabp4 mt-Cytb 0.8 0.5 Uqcrfs1 0.8 Acad11 Retsat Acat1 Ndufb8 Tmem120a Ndufb9 Abcd2 Acaa1a Sfxn1 Ndufb11 0.6 Aldh1l1 Cox4i1 Adipoq Acadvl Adipor2 Cyc1 Cidec Sdhd Auh Ak2 Park7 Acadm Hsd17b4 Ndufs1 Cd36 Acaa2 Ndufs2 Prkar2b Uqcrc1 Sort1 Sdhb Mknk2 Ndufv1 Lpl Cox7b d Macrophage Hadhb Higd2a Hadha Cox17 & angiogenesis Tspan12 markergenes Mgll c Chemokine genes Rgcc WT KO Cs WT KO GC FL GC FL Cdo1 Por GC FL GC FL Cd68 Vegfa Ccl2 Lgals3 Cebpa Ccl7 6 Cav2 Kdr 1.5 Gnpat Ccl3 4 G0s2 Ccl5 Pecam1 Mrap Lpin1 Ccl11 2 Adgre1 1.0 Scd1 Ccl8 Thrsp Fasn f WAT Space stresses Nrf2 deficiency WT KO

Inflammation GC Down -regulation of PPAR Angiogenesis Respiratory chain Fatty acid -oxidation Adipocyte number FL Metabolic impairment Compensatory enlargement

Enlargement of adipocytes Impaired response to space stress in response to space stresses

Fig. 7 Space stresses and Nrf2-deficiency induce metabolic impairment in eWAT. a–e Heatmap of expression levels of genes involving in the respiratory chain (a), fatty acid β-oxidation (b), chemokine (c), macrophage and angiogenesis (d) and downstream of PPARγ (e) in eWAT from WT and Nrf2-KO mice in GC and FL groups. f Hypothetical model for space-induced metabolic impairment in eWAT from WT and Nrf2-KO mice in GC and FL groups. of oxidative stress. Similarly, there is a possibility that mechanical returned safely to Earth. Comprehensive RNA sequencing ana- stress caused by microgravity may contribute to Nrf2 activation lyses of various organs/tissues from the returned mice revealed during space travel. However, there has been limited direct that space stresses indeed have induced the Nrf2 signaling examination as to whether these space stresses activate the pathway in many tissues. Our metabolome analysis further Nrf2 signaling pathway and to what extent activation of the revealed that the stresses during the space travel have induced pathway may be protective against these stresses. Therefore, we ageing-related changes of some plasma metabolites. Another conducted a space travel experiment utilizing Nrf2 knockout salient finding in the space travel of Nrf2 knockout mice is that mice. To our best knowledge, this is the first attempt of space Nrf2 is important for maintaining homeostasis of white adipose travel for gene-knockout disease model mice in which mice are tissues. Collectively, these results thus unequivocally demonstrate

COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio 11 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01227-2 the importance of the Nrf2 signaling pathway in responses to the our series of histological analyses as Supplementary Data (Sup- stresses of space travel. plementary Figs. 11–14). Further metabolomic analyses, blood Many technological advances contributed to the success of this analyses, kidney function analyses, behavioral analyses, and space mission, including the mouse transport and habitat systems. muscle analyses, will be published separately. Collectively, this Of direct application to the biomedical inquiries, we developed a study has pioneered the future of experiments utilizing gene- new blood collection procedure that is minimally invasive and easy- modified model mouse defective in adaptive responses to stresses for-use for astronauts. This apparatus is important since there is a associated with space travel. We believe that continuation of this limitation for the training of astronauts for specialized operations series of space mouse studies will provide insightful information such as blood collection. Capitalizing on the development of the useful to overcome possible mission-limiting, space flight-derived new device, we successfully collected small amounts (40 μL) of stresses evoked by human space exploration. blood and plasma from mouse tails during the flight. Subsequent mass spectrometer-based metabolome analysis successfully detected Methods plasma metabolites in the small volumes of plasma. This newly The MHU-3 project. The HCU rev.1 is an onboard habitation cage that accom- established procedure of blood collection from mice during space modates one mouse per cage11,12. The HCU is equipped with a food bar, watering travel coupled with a highly sensitive metabolome analysis brings us system (two redundant water nozzles and a water balloon acting as power-free pressure source), an odor filter, two fans (for ) for air ventilation, waste a powerful approach to elucidate physiological and pathological collecting equipment, an LED/IR video camera with a wiper inside the cage to keep changes of intermediary metabolism in gene-modified or other the observation window clean. The health of each mouse was determined based on model mice during space travel. the conditions of eyes, ears, fur, and tail as observed in the transmitted videos. Indeed, the systematic metabolome analysis in this study Paper sheets were mounted on the cage wall to quickly eliminate liquid, such as fl urine, from the cage. A photocatalytic thermal spray was applied to the sheets for revealed that space ight induced increases in glycerol and deodorant and antibacterial effects. Air ventilation inside the cage was maintained decreases in TG, implicating the occurrence of enhanced lipolysis by airflow (<0.2 m/s) generated by fans on the HCU. Differences in the volume of in mice during the flight. It has been shown that lipolysis is air ventilation and airflow rates between HCU in GC and microgravity conditions stimulated by stress-induced catecholamines, including adrena- were negligible because the fans regulating ventilation in both gravity conditions 27 were maintained at the same speed. The day/night cycle used 12-h intervals. Food line and noradrenaline , and an increase in circulating levels of and water were replenished once a week. Temperature, humidity, carbon dioxide, 28 catecholamines has been observed commonly in astronauts . and ammonia were monitored precisely and recorded in logs. Therefore, we surmise that the enhanced lipolysis is the con- The TCU was used to transport mice aboard the SpaceX Dragon capsule during sequence of elevated catecholamine hormones. Alternatively, the the launch and return phases, and was placed in a powered locker sized for ISS single cargo transfer bag. The TCU contains 12 cylindrical cages for housing mice enhanced lipolysis in mice might be a compensatory response to individually. The TCU is equipped with a cylindrical food bar, watering system (a the increase in adipose mass occurring during space travel. In this water nozzle and two water balloons), an odor filter, two fans (for redundancy), regard, it is interesting to note that the increase of plasma glycerol waste collecting equipment, and LEDs with lights for day/night cycle. Paper sheets shows significant association with human ageing as observed in mounted in the waste collection area were treated with photocatalytic thermal the ToMMo cohort study. Now, verification of this relationship in spray for deodorant and antibacterial effects, similar to the HCU. A temperature/ humidity logger was attached on the TCU air inlet to monitor the environment prospective cohort studies becomes quite important and intri- during launch and reentry operations. guing, since association of the glycerol increase and ageing has not been recognized heretofore. Animals. Nrf2-KO (Nfe212tm1Ymk)16 and WT male mice in the C57BL/6J back- We also observed that decreased plasma glycine levels showed ground were bred at Charles River Laboratories Japan for MHU3. All animal a marked association with ageing. In contrast to the situation for experiments were approved by the Institutional Animal Care and Use Committees glycerol, the association of plasma glycine levels with ageing has of JAXA (protocol numbers 017-001 and 017-014), NASA (protocol number FLT- 17-112), and Explora BioLabs (EB15-010C), and conducted according to the been reported, such that dietary glycine supplementation exten- related guidelines and applicable laws of Japan and the United States of America. ded the lifespan of rats29 and addition of glycine to the culture media restored the phenotype of aged cells back to that of young Pre-launch acclimation activities and animal selection. Three weeks prior to 30 cells . Taken together, our results and these reports indicate that launch, 60 WT and 60 Nrf2-KO mice (8 weeks old) were delivered from Charles space travel induces in mice an ageing phenotype associated with River Laboratories Japan to KSC. These mice were acclimatized to the environment in individual housing cages (Small Mouse Isolator 10027, Lab Products) in an air- these plasma metabolites. – Our finding that Nrf2 disruption impairs body-weight gain and conditioned room (temperature: 23 ± 3 °C; humidity: 40 65%) with a 12:12-h light/ fl dark cycle at the SSPF Science Annex at KSC. Three acclimation phases were set: WAT homeostasis in space mice during ight elicits several Phase I, body weight recovery phase; Phase II, water nozzle acclimation phase and important considerations. First, it has been known that there is a Phase III, flight food acclimation phase. After approximately three weeks of specific regulatory single nucleotide polymorphism (rSNP) in the acclimation, the mice were ready for transport to the ISS and had reached the age Nrf2 gene that downregulates the expression of Nrf2 and increases of 12 weeks. Succinctly, at Phase I, mice were fed CRF-1 and given autoclaved tap 31 32 water ad libitum using ball-type water nozzles. The bedding material consisted of the risk of a few diseases in humans and mice . Thus, a number paper chips (ALPHA-DRI). At Phase II, water nozzles are changed to flight nozzles. of questions arise related to this rSNP. For instance, does presence Finally, at Phase III, the food was changed to flight food. Fecal collection and body of this rSNP in the NRF2 gene influence the health of future longer- swabs were obtained, and these samples were sent to Charles River Laboratories in term space travelers? Do minor allele homozygote mice of the rSNP USA for pol-based SPF testing. As some of the Nrf2-KO mice harbor an intrahepatic shunt35, Nrf2-KO mice retain exacerbated risk for body-weight reduction and/or pertur- were challenged with 100 mg/kg ketamine and 5 mg/kg xylazine and recovery times bation of WAT homeostasis by long-term space flight or by ageing? were monitored at the age of 9 weeks to eliminate the mice harboring an Second, small molecule inducers of the Nrf2 signaling have been intrahepatic shunt. Nrf2-KO mice with long sleeping duration were judged as mice developed or are under development33,34. These NRF2 inducers can without shunts. Blood was taken from tails of all 120 mice. During the acclimation phase, body weight and food/water consumption were be taken as drugs or through foods and dietary supplements. We measured and recorded. We selected 12 mice for flight and backup, respectively. surmise that these NRF2 inducers may serve to mitigate some of the The selection of flight candidate mice was based on body weight, food stresses associated with space travel. consumption, water intake, and absence of hepatic shunt. We also checked the We have obtained substantive amounts of informative data on health of each flight candidate mouse by evaluating the condition of their eyes, ears, –fl teeth, fur, and tail. The five subgroups of flight candidates were necessary to this MHU-3 space ight study. However, we could analyze only support possible launch attempts in case of unexpected postponements. limited sets of data in a timely manner. In fact, we feel that many changes were induced in a space–flight specific manner and/or fi Onboard operations by ISS crew. One day prior to launch, 12 mice were loaded Nrf2-knockout mouse speci c manner that have not been into the TCU and made ready for launch. Mice were transported to the ISS by recognized through our first-pass analyses. We therefore present SpX14. After the Dragon vehicle of SpX14 berthed with the ISS, mice were

12 COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01227-2 ARTICLE transferred to the HCU by a crew member. Mouse-husbandry tasks included was used to assess space flight-induced changes of gene expression. The GSEA was exchanging food cartridges, supplying water, collecting waste, and replacing odor performed using previously published ageing signatures in Enrichr40 (GSE20425 for filters. The block food (approximately 35 g) is contained in the cartridge, which liver, GDS3028 for temporal bone, GSE25325 for BAT and GSE25905 for WAT) or 50 supports 1-week habitation. The cartridges have windows and scale at their sides, hallmarked gene sets of MSigDB v7.1 (www.gsea-msigdb.org/gsea/msigdb/index.jsp). therefore the remaining amount of the foods was estimated, via video without returning them to the ground, at weekly food cartridge exchange operations. After Histological analysis. Tissues were fixed in Mildform 10N (Wako Pure Chemical) 31 days onboard the ISS, mice were transferred to the TCU for the return to Earth. and processed into paraffin-embedded tissue sections. The sections were stained with hematoxylin and eosin. Lipid droplet size was measured by using BZ-X800 Return phase and animal dissection. After unberthing from the ISS, the Dragon (Keyence), and adipose cell number was calculated from their tissue weights and vehicle splashed down in the Pacific Ocean off the coast of California. After a ship droplet sizes41,42. To visualize hepatic lipid content, livers were fixed with 4% picked up the cargo, the returned TCU was transported to a port in Long Beach. paraformaldehyde and embedded in OCT (Tissue Tek). The frozen sections were JAXA received the TCU from NASA and transported them to Explora BioLabs in stained with Oil Red O (Sigma Aldrich) and counterstained with hematoxylin. San Diego using an environmentally controlled van. After the health and body weights of mice were assessed, open field, light/dark transition and Y-maze tests fl Plasma analysis by NMR spectroscopy. Plasma metabolites of the blood samples were performed. Blood was collected from tail after the behavioral tests. Iso urane- obtained from inferior vena cava were analyzed using NMR spectroscopy17,18. anesthetized mice were then euthanized by exsanguination and dissected for the Plasma metabolites were firstly extracted using a standard methanol extraction collection of tissue samples. procedure using 50 µL of plasma per sample. The supernatant was transferred to a new tube and evaporated. Each dried sample was suspended in a 200-µL of GC experiment. A GC experiment that simulated the space experiment was con- 100-mM sodium phosphate buffer (pH 7.4) in 100% D2O containing 200-µM d6- ducted at JAXA Tsukuba in Japan from September 17 to October 20, 2018. Six WT DSS. NMR experiments were performed at 298 K on a Bruker Avance 600-MHz and six Nrf2-KO mice without the hepatic shunts were individually housed in the spectrometer equipped with a CryoProbe and a SampleJet sample changer. Stan- same way as for the flight experiment. Both the TCU and HCU were placed in an air- dard 1D NOESY and CPMG (Carr-Purcell-Meiboom-Gill) spectra were obtained conditioned room (average temperature: 22.9 °C; average humidity: 48.4%) with a for each plasma sample. All data were processed using the Chenomx NMR Suite 12:12-h light/dark cycle. Fan-generated airflow (0.2 m/s) inside the HCU maintained 8.3 processor module (Chenomx). Metabolites were identified and quantified using the same conditions as in the flight experiment. The mice were fed CRF-1. The the target profiling approach implemented in the Chenomx Profiler module. The bedding material consisted of paper chips (ALPHA-DRI) in the acclimation cages, but of metabolites was analyzed by a PCA on SIMCA13.0.0 (Umetrics). was not used in the HCU. Feeder cartridges and water bottles were replaced once a week, and cages in the HCU were not replenished. After the health and body weights fi Plasma analysis by mass spectrometry. The plasma from tail blood sample was of mice were assessed, open eld, light/dark transition and Y-maze tests were per- collected from the capillary tube and stored at −80 °C until analysis. Plasma (4 μL formed. Blood sampling from the tail was conducted as for the flight experiment. All ® fl per analysis) were prepared using the protocol for the Absolute IDQ p400 HR Kit mice were anesthetized by iso urane inhalation and euthanized under anesthesia, and (Kit400), which includes a detailed standard operating procedure (SOP) protocol then dissected to collect tissue samples. with documentation for sample preparation, instrument setup, system suitability testing, and data analysis. The Kit400 quantifies 408 metabolites, and includes Blood collection procedure. During the flight experiment, blood samples were calibration standards, internal standards, and quality control samples. The ultra- collected from the distal end of the tail with tail clippers (KAI, PQ3357). The tail high-performance liquid chromatography (UHPLC) system consisted of an online clippers were mounted on a guard plate adjusted to 1 mm from the tip of the degasser, auto sampler, dual pump, and column oven (UltiMateTM 3000 RSLC clippers, and all such tail clippers used in this experiment were inspected to con- system, Thermo Fisher Scientific), and a quadrupole Fourier transform mass firm each tool’s ability to amputate less than 1 mm of the tail prior to launch. spectrometry (FTMS, Q Exactive Orbitrap system). The operating conditions of the Each mouse was transferred from the HCU to the restrainer (Sanplatec, 99966-31), UHPLC-FTMS system and the details of data analysis followed the protocol of the which was pre-warmed by disposable heat pads (Kiribai Chemical; New Hand Warmer Kit400 analysis43. Mini)toincreaseobtainablebloodvolume.Theinnersurfacetemperatureofthe restrainer was approximately 45 °C. After being disinfected, the tail was milked after Statistical and reproducibility. Data points represent biological replicates. being clipped by a tail clipper to collect a blood sample into capillary tubes (Drummond fi Comparisons between groups were conducted using one-way ANOVA with Scienti c; 8-000-7520-H/5). To promote hemostasis, the tip of mouse tail was Tukey–Kramer test or Wilcoxon–Mann–Whitney test. Data were considered sta- compressed for one minute with a hemostat (Ethicon, 15726). The blood samples tistically significant at p < 0.05. collected in capillary tubes were centrifuged and immediately frozen in a −80 °C-freezer without being snap-frozen. Prior to blood collection, a veterinarian checked the health status of all mice via videos and confirmed their adaptation to the space environment. Reporting summary. Further information on research design is available in the Nature Before and after the flight experiment, blood samples were taken by nicking the Research Reporting Summary linked to this article. tail. Mice were transferred to the restrainers and then nicked using disposable scalpels (FEATHER Safety Razor, No. 10). After the nicking or the on-orbit Data availability procedure, blood was collected in the capillary tubes and centrifuging was The data discussed in this publication have been deposited in NCBI’s Gene Expression conducted. The procedures for both nicking and clipping were in accordance with 44 NIH guidelines. Omnibus and are accessible through GEO Series accession number GSE152382 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE152382). All relevant data are available from the corresponding authors upon reasonable request. Supplementary Micro-computed tomography (microCT) analysis. The right femurs of mice Data 1 contains gene set enrichment analysis (GSEA) of the gene expression changes fi 11 were xed in 70% ethanol and distal regions were analyzed as described . shown during the space flight in wild-type mice. Source data for graphs and charts MicroCT scanning was performed using a ScanXmate-A100S Scanner (Coms- presented in the main figures are provided in Supplementary Data 2. cantechno). Three-dimensional microstructural image data was reconstructed and BMD was calculated using TRI/3D-BON software (RATOC System Engineering) in accordance with the guidelines. Received: 15 April 2020; Accepted: 13 August 2020;

RNA-sequence analysis. Total RNA was isolated from temporal bone, mandibular bone, spleen, liver, white adipose tissue, brown adipose tissue, cerebrum, kidney, and thymus. RNA integrity was assessed using an Agilent 2200 TapeStation. For each tissue, 1.0 μg of total RNA was used for further steps. Total RNA samples were subjected to isolation of poly(A)-tailed RNA and library construction using Sureselect References fi Strand Speci c RNA Sample Prep Kit (Agilent Technologies), except that total RNA 1. Demontis, G. C. et al. Human pathophysiological adaptations to the space from thymus was subjected to ribosomal RNA depletion using Ribo-Zero rRNA environment. Front. Physiol. 8, 547 (2017). Removal Kit (Illumina) and library construction using the similar step. The libraries 2. Pecaut, M. J. et al. Genetic models in applied physiology: selected contribution: were sequenced using HiSeq2500 (Illumina) for 76 cycles of single read and more than effects of spaceflight on immunity in the C57BL/6 mouse. I. Immune 17 million reads were generated per sample. The raw reads were mapped to the mouse population distributions. J. Appl. Physiol. 94, 2085–2094 (2003). mm10 genome using STAR (version 2.6.1)36. Transcripts per million (TPM) values 3. Baqai, F. P. et al. Effects of spaceflight on innate immune function and were obtained to measure gene expression using RSEM (version 1.3.1)37.TheTPM antioxidant gene expression. J. Appl. Physiol. 106, 1935–1942 (2009). was normalized in the Subio Platform software (Subio). Genes with mean TPM less fl fi 4. Behnke, B. J. et al. Effects of space ight and ground recovery on mesenteric than ve among all groups were excluded. The PCA was performed using the prcomp – function in R software (version 3.5.1; www.r-project.org). R-based heatmap.2 in gplots artery and vein constrictor properties in mice. FASEB J. 27, 399 409 (2013). package was used for generating the heatmap. The differential gene expression analysis 5. Gridley, D. S. et al. Changes in mouse thymus and spleen after return from the was performed using DESeq2 (version 1.22.2)38. GSEA39 (www.broad.mit.edu/gsea) STS-135 mission in space. PLoS ONE 8, e75097 (2013).

COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio 13 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01227-2

6. Andreev-Andrievskiy, A. et al. Mice in Bion-M 1 space mission: training and 39. Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based selection. PLoS ONE 9, e104830 (2014). approach for interpreting genome-wide expression profiles. Proc. Natl Acad. 7. Ulanova, A. et al. Isoform composition and gene expression of thick and thin Sci. USA 102, 15545–15550 (2005). filament in striated muscles of mice after 30-day space flight. Biomed. 40. Kuleshov, M. V. et al. Enrichr: a comprehensive gene set enrichment analysis Res. Int. 2015, 104735 (2015). web server 2016 update. Nucleic Acids Res. 44, W90–W97 (2016). 8. Cancedda, R. et al. The Mice Drawer System (MDS) experiment and the space 41. Parlee, S. D., Lentz, S. I., Mori, H. & MacDougald, O. A. Quantifying size and endurance record-breaking mice. PLoS ONE 7, e32243 (2012). number of adipocytes in adipose tissue. Methods Enzymol. 537,93–122 (2014). 9. Tavella, S. et al. Bone turnover in wild type and pleiotrophin-transgenic mice 42. Wang, J. et al. Phytol increases adipocyte number and glucose tolerance housed for three months in the International Space Station (ISS). PLoS ONE 7, through activation of PI3K/Akt signaling pathway in mice fed high-fat and e33179 (2012). high-fructose diet. Biochem. Biophys. Res. Commun. 489, 432–438 (2017). 10. Sandonà, D. et al. Adaptation of mouse skeletal muscle to long-term 43. Thompson, J. W. et al. International Ring Trial of a high resolution targeted microgravity in the MDS mission. PLoS ONE 7, e33232 (2012). metabolomics and lipidomics platform for serum and plasma analysis. Anal. 11. Shiba, D. et al. Development of new experimental platform ‘MARS’-Multiple Chem. 91, 14407–14416 (2019). Artificial-gravity Research System-to elucidate the impacts of micro/partial 44. Edgar, R., Domrachev, M. & Lash, A. E. Gene Expression Omnibus: NCBI gravity on mice. Sci. Rep. 7, 10837 (2017). gene expression and hybridization array data repository. Nucleic Acids Res. 30, 12. Matsuda, C. et al. Dietary intervention of mice using an improved Multiple 207–210 (2002). Artificial-gravity Research System (MARS) under artificial 1. NPJ Microgravity 5, 16 (2019). 13. Yamamoto, M., Kensler, T. W. & Motohashi, H. The KEAP1-NRF2 System: a Acknowledgements thiol-based sensor-effector apparatus for maintaining redox homeostasis. We would like to thank Norishige Kanai (astronaut) for the onboard operation, and Physiol. Rev. 98, 1169–1203 (2018). Toshiaki Kokubo and Noriko Kajiwara (JAXA visiting veterinarians) for monitoring 14. Hosoya, T. et al. Differential responses of the Nrf2-Keap1 system to laminar mouse health. We also thank Naoko Ota-Murakami, Fumika Yamaguchi, Masumi and oscillatory shear stresses in endothelial cells. J. Biol. Chem. 280, Umehara, and the members of the mouse health check team, for performing daily 27244–27250 (2005). onboard health checks, Ramona Bober, Autumn L. Cdebaca, Rebecca A. Smith for 15. Indo, H. P. et al. Changes in mitochondrial homeostasis and redox status in animal care and ground experiment supports, Hirochika Murase, Hiroaki Kodama, astronauts following long stays in space. Sci. Rep. 6, 39015 (2016). Yusuke Hagiwara, and members of hardware development team for MHU hardware 16. Itoh, K. et al. An Nrf2/small Maf heterodimer mediates the induction of phase preparation and operations, Kohei Hirakawa, Teruhiro Senkoji, Haruna Tanii, Motoki II detoxifying genes through antioxidant response elements. Biochem Tada, Yuki Watanabe, Kayoko Lino, Hiromi Sano, Yui Nakata, Hiromi Suzuki-Hashi- Biophys. Res. Commun. 236, 313–322 (1997). zume, Eiji Ohta, Osamu Funatsu, Hideaki Hotta, Hatsumi Ishida, Mariko Shimizu, and 17. Koshiba, S. et al. The structural origin of metabolic quantitative diversity. Sci. members of JEM operational team for the research coordination, Takahashi Ueda and Rep. 6, 31463 (2016). Tomohiro Tamari for animal preparations, Hong Xin and Grishma Acharya for landing 18. Koshiba, S. et al. Omics research project on prospective cohort studies from site operational supports, and Sayaka Umemura, Laura Lewis, Charles E. Hopper, Jen- the Tohoku Medical Megabank Project. Genes Cells 23, 406–417 (2018). nifer J. Scott Williams, Robert Kuczajda for international coordination. This work was 19. Saigusa, D. et al. Establishment of protocols for global metabolomics by LC- selected as a space rodent research study for JAXA’s feasibility experiments using ISS/ MS for biomarker discovery. PLoS ONE 11, e0160555 (2016). Kibo announced in 2015, and also supported in part by MEXT/JSPS KAKENHI 20. Kim, J. W. et al. Characterizing genomic alterations in cancer by (19H05649 to M.Y. and 17KK0183, 18H04963, 19K07340, to T. Suzuki), Takeda Science complementary functional associations. Nat. Biotechnol. 34, 539–546 (2016). Foundation (M.Y., and T. Suzuki.) and the Smart Aging Research Center, Tohoku 21. Tominari, T. et al. Hypergravity and microgravity exhibited reversal effects on University (M.Y.). This work was also supported in part by the grants JP19km0105001, the bone and muscle mass in mice. Sci. Rep. 9, 6614 (2019). JP19km0105002 and SHARE. 22. Tadaka, S. et al. jMorp: Japanese multi omics reference panel. Nucleic Acids Res. 46, D551–D557 (2018). Author contributions 23. Huber, J. D. Diabetes, cognitive function, and the blood-brain barrier. Curr. T.S., A.U., A.Y., T.W.K., D.S., and M.Y. wrote the paper. A.Y., M. Shimomura, H. Pharm. Des. 14, 1594–1600 (2008). Mizuno, M. Shirakawa, and D.S. conducted the space experiments. T.S., A.U., A.Y., K.T., 24. Tahergorabi, Z. & Khazaei, M. The relationship between inflammatory M. Suzuki, N.H., R.R., E.N., N.O., A.G., H.S., R.B., A.O., F.K., T.Y., D.S., S.K., T.N., S.F., markers, angiogenesis, and obesity. ARYA Atheroscler. 9, 247–253 (2013). H.I., K.N., N.S., I.H., R.S., T.O., H. Motohashi, H.T., R.O., T.K., and S.T. conducted the 25. Lehrke, M. & Lazar, M. A. The many faces of PPARγ. Cell 123,993–999 (2005). molecular, histological, physiological, and behavioral assessments and analyzed the data. 26. Suzuki, T. et al. Molecular mechanism of cellular oxidative stress sensing by Keap1. Cell Rep. 28, 746–758.e744 (2019). 27. Lönnqvist, F., Nyberg, B., Wahrenberg, H. & Arner, P. Catecholamine- Competing interests induced lipolysis in adipose tissue of the elderly. J. Clin. Investig. 85, The authors declare no competing interests. 1614–1621 (1990). 28. Weil-Malherbe, H., Smith, E. R. & Bowles, G. R. Excretion of catecholamines Additional information and catecholamine metabolites in Project Mercury pilots. J. Appl. Physiol. 24, Supplementary information 146–151 (1968). is available for this paper at https://doi.org/10.1038/s42003- 29. Miller, R. A. et al. Glycine supplementation extends lifespan of male and 020-01227-2. female mice. Aging Cell 18, e12953 (2019). Correspondence 30. Hashizume, O. et al. Epigenetic regulation of the nuclear-coded GCAT and and requests for materials should be addressed to D.S. or M.Y. SHMT2 genes confers human age-associated mitochondrial Reprints and permission information defects. Sci. Rep. 5, 10434 (2015). is available at http://www.nature.com/reprints 31. Suzuki, T. et al. Regulatory nexus of synthesis and degradation deciphers – Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in cellular Nrf2 expression levels. Mol. Cell Biol. 33, 2402 2412 (2013). fi 32. Cho, H. Y. et al. Linkage analysis of susceptibility to . Nrf2 is a published maps and institutional af liations. candidate gene. Am. J. Respir. Cell Mol. Biol. 26,42–51 (2002). 33. Cuadrado, A. et al. Therapeutic targeting of the NRF2 and KEAP1 partnership in chronic diseases. Nat. Rev. Drug Discov. 18, 295–317 (2019). Open Access This article is licensed under a Creative Commons 34. Yagishita, Y., Fahey, J. W., Dinkova-Kostova, A. T. & Kensler, T. W. Broccoli Attribution 4.0 International License, which permits use, sharing, or sulforaphane: is it the source or dose that matters? Molecules https://doi. adaptation, distribution and reproduction in any medium or format, as long as you give org/10.3390/molecules24193593 (2019). appropriate credit to the original author(s) and the source, provide a link to the Creative 35. Skoko, J. J. et al. Loss of Nrf2 in mice evokes a congenital intrahepatic shunt Commons license, and indicate if changes were made. The images or other third party that alters hepatic oxygen and protein expression gradients and toxicity. material in this article are included in the article’s Creative Commons license, unless Toxicol. Sci. 141, 112–119 (2014). indicated otherwise in a credit line to the material. If material is not included in the 36. Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, article’s Creative Commons license and your intended use is not permitted by statutory 15–21 (2013). regulation or exceeds the permitted use, you will need to obtain permission directly from 37. Li, B. & Dewey, C. N. RSEM: accurate transcript quantification from the copyright holder. To view a copy of this license, visit http://creativecommons.org/ RNA-Seq data with or without a reference genome. BMC Bioinform. 12, 323 licenses/by/4.0/. (2011). 38. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014). © The Author(s) 2020

14 COMMUNICATIONS BIOLOGY | (2020) 3:496 | https://doi.org/10.1038/s42003-020-01227-2 | www.nature.com/commsbio